In my opinion, Appian RPA can be improved. We are planning to conduct a proof of concept on volume because we are still not entirely certain about how well the record type feature performs on Appian with high volumes of data. After conducting client calls with HSBC, Santander, and a few of the Big Four, we learned about use cases where Appian has been used for high volumes of data. We are planning to conduct a proof of concept to test that hypothesis. Additional features that could enhance Appian RPA in the future would include AI on Azure. Currently, their AI is only AWS-based, and we are in discussions with the company about this because we are an Azure shop. I believe this would be a good area for them to concentrate on.
Director & Key Managerial Personnel at Newlight Technology Solutions Private Limited
Real User
Top 5
Jul 23, 2025
There are many competitive products in the market right now, for example, UiPath and Automation Anywhere. These competitive products are now coming up with features which Appian RPA currently does not have. For example, agentic AI automation is a new feature which UiPath recently released, and while Appian is planning to release something similar soon, it is not at the same level as UiPath yet. Appian could work more on the agentic AI automation side of things for RPA.
There are not any areas that need improvement off the top of my head. It has been 18 months since I worked with the environment, and they've made UI enhancements and implemented AI, including features like Copilot.
Process Automation streamlines operations by reducing manual intervention, enhancing efficiency, and ensuring task consistency. It integrates seamlessly into existing workflows to optimize productivity.Businesses utilize Process Automation to achieve high levels of accuracy in repetitive tasks. Automation solutions decrease workload pressure, allowing employees to focus on higher-value activities. This technology ensures quicker processing times and enhances overall performance within diverse...
In my opinion, Appian RPA can be improved. We are planning to conduct a proof of concept on volume because we are still not entirely certain about how well the record type feature performs on Appian with high volumes of data. After conducting client calls with HSBC, Santander, and a few of the Big Four, we learned about use cases where Appian has been used for high volumes of data. We are planning to conduct a proof of concept to test that hypothesis. Additional features that could enhance Appian RPA in the future would include AI on Azure. Currently, their AI is only AWS-based, and we are in discussions with the company about this because we are an Azure shop. I believe this would be a good area for them to concentrate on.
There are many competitive products in the market right now, for example, UiPath and Automation Anywhere. These competitive products are now coming up with features which Appian RPA currently does not have. For example, agentic AI automation is a new feature which UiPath recently released, and while Appian is planning to release something similar soon, it is not at the same level as UiPath yet. Appian could work more on the agentic AI automation side of things for RPA.
There are not any areas that need improvement off the top of my head. It has been 18 months since I worked with the environment, and they've made UI enhancements and implemented AI, including features like Copilot.